The design of buildings has become a predominantly digital affair, with computers replacing drafting boards. The recent surge in Artificial Intelligence (AI) has only fueled this transition. However, the importance and relevance of analog-making persist in sketching. Despite the widespread adoption of generative AI, particularly text-to-image platforms like Stable Diffusion, Midjourney, and DALL·E, sketching remains indispensable in both analog and digital formats. Generative AI tools complement and enhance many qualities that proponents of sketching value. Andrew Kudless points out that “AI models are inherently sketchy—imperfect and incomplete works embedded with both opportunities and challenges” (Kudless 2024, 93). Sketches are imprecise, loose, and abstract. Like sketches, generative AI tools can create abstract images that aid decision-making without overly defining the outcome. However, despite the continued use of sketching as a design tool, it remains largely confined to a single domain—you sketch by hand or on a computer. Furthermore, much has been written about AI as a sketch tool, but less about the trajectory of this human-machine exchange and the logical next step in the analog-digital process: collage-making.
